PPL
Bernhard Schölkopf
Papers (6)
PEFT-Arena: Understanding Parameter-Efficient Finetuning from a Stability-Plasticity PerspectiveLearning to Reason Efficiently with A* Post-TrainingRiemannian Networks over Full-Rank Correlation MatricesMechanism Design Is Not Enough: Prosocial Agents for Cooperative AIAdaptive Inverted-Index Routing for Granular Mixtures-of-ExpertsNeural Posterior Estimation of Terrain Parameters from Radar Sounder Data